1. Field of the Invention
The present invention relates to an image recognition device that recognizes a position of a recognition object in an image captured by a camera, an image recognition method, an image recognition program, and a recording medium.
2. Description of the Related Art
In recent years, have been proposed many image recognition technologies for recognizing a position of a predetermined recognition object in an image captured by a camera. In order to perform this position recognition with high accuracy, it is necessary to take sufficient measures to perform correction of a captured image, etc. with respect to a vibration applied to the camera is subjected.
In contrast with this, for example, a prior art described in JP, A, 2006-317848 detects a movement vector of a background of an image captured by a camera, and performs correction to user's camera shake by moving an optical-axis shift lens by means of a vector opposite to the movement vector, thereby allowing for panning without blurring of a subject.
Meanwhile, in recent years, as an operation input method for an interface device mounted on a moving body, such as a vehicle, for example, has been proposed a gesture operation input technology for performing image recognition of a position of a hand raised by a user in front of a camera included in the interface device, and inputting a selection indication corresponding to the position. Here, although a vibration (blurring) is generated also at the position of the hand which is image-recognized due to generation of a vibration associated with running of the moving body, a relative vibration between a body itself of the user and the moving body is also synthesized and included in the vibration in this case in addition to a relative vibration between the interface device itself including the camera and the moving body.
In the above-described prior art, although correction to the relative vibration between the camera itself and the moving body can be performed, the relative vibration between the body itself of the user and the moving body cannot be dealt with at all. For this reason, there has been demanded a technology that can improve recognition accuracy in recognizing a position of a predetermined body portion of the user riding on the moving body by means of an image captured by the camera mounted on the moving body.